Background Regular exercise is associated with enhanced nitric oxide (NO) bioavailability. Flow-mediated dilation (FMD) is used widely to assess endothelial function (EF) and NO release. Objectives The aims of this systematic review and metaanalysis were to (i) investigate the effect of exercise modalities (aerobic, resistance or combined) on FMD; and (ii) determine which exercise and participant characteristics are most effective in improving FMD. Methods We searched the MEDLINE, Embase, Cochrane Library, and Scopus databases for studies that met the following criteria: (i) randomized controlled trials of exercise with comparative non-exercise, usual care or sedentary groups; (ii) duration of exercise intervention C4 weeks; (iii) age C18 years; and (iv) EF measured by FMD before and after the intervention. Weighted mean differences (WMDs) with 95 % confidence interval were entered into a random effect model to estimate the pooled effect of the exercise interventions. Results All exercise modalities enhanced EF significantly: aerobic (WMD 2.79, 95 % CI 2.12–3.45, p = 0.0001), resistance (WMD 2.52, 95 % CI 1.11–3.93, p = 0.0001) and combined (WMD 2.07, 95 % CI 0.70–3.44, p = 0.003). A dose–response relationship was observed between aerobic exercise intensity and improvement in EF. A 2 metabolic equivalents (MET) increase in absolute exercise intensity or a 10 % increase in relative exercise intensity resulted in a 1 % unit improvement in FMD. There was a positive relationship between frequency of resistance exercise sessions and improvement in EF (b 1.14, CI 0.16–2.12, p = 0.027). Conclusions All exercise modalities improve EF signifi- cantly and there was a significant, positive relationship between aerobic exercise intensity and EF. Greater frequency, rather than intensity, of resistance exercise training enhanced EF.